Large language models (LLMs) are limited by complex reasoning tasks that require multiple steps, domain-specific knowledge, or external tool integration. To address these challenges, researchers have ...